Manufacturers see AI gains but data gaps stall scale-up

Manufacturers are reporting early financial returns from AI projects, but most organisations still lack the readiness and data confidence needed to deploy the technology at scale, according to Riverbed research.

Riverbed's survey of manufacturing business and IT leaders and technical specialists found that 87% said return on investment from AIOps initiatives met or exceeded expectations. Even so, 62% of AI projects remained in pilot or development, and only 37% said they were fully prepared to operationalise AI at scale.

The results highlight a widening gap between experimentation and deployment. Many manufacturers have pushed AI into frontline use cases such as predictive maintenance and troubleshooting, which depend on high-quality operational data and reliable industrial networks.

Data confidence

Data quality emerged as a key constraint. Nearly half of respondents (47%) said they lacked confidence in the accuracy and completeness of their organisation's data. Only 34% rated their data as excellent for relevance and suitability.

At the same time, 90% agreed that improving data quality is critical to AI success. Riverbed also found that 57% of manufacturing organisations expressed confidence in their AI projects, suggesting a split between leadership optimism and operational reality.

Richard Tworek, Riverbed's chief technology officer, said manufacturers are investing heavily but still face obstacles as they move beyond early projects.

"The manufacturing industry is investing heavily in AI to transform IT operations, and our survey results show that nearly nine in ten companies in this sector (87%) are already meeting or exceeding ROI expectations from their AIOps investments," Tworek said.

"However, many still face major challenges, including gaps in readiness and preparedness, as well as data quality issues that are hindering progress. As a data-driven company, we're helping our manufacturing customers close these gaps with safe, secure and accurate AI built on high-quality real data, delivering practical AI-powered solutions that enable organisations to scale AI across the enterprise," he added.

Tool sprawl

The research suggests IT operations teams are managing complex tool estates that can undermine consistent data practices. Manufacturing organisations reported using an average of 13 observability tools from nine vendors.

In response, 95% said they were consolidating tools, and 91% said they were considering new tools as part of those plans. The main drivers were improved integration and interoperability (48%), reduced vendor-management overhead (47%), and improved IT productivity (46%).

Consolidation often intersects with AI deployment because data collection and correlation across systems affect model performance and the reliability of automated actions. The results suggest many organisations see rationalisation as a necessary step towards more consistent operational insight.

Workplace systems

Manufacturers also flagged problems with unified communications tools, which have become more important as organisations mix on-site work with remote teams and external partners. The survey found that 42% of employees used unified communications tools throughout their work week, while 66% of respondents said they were essential to operating effectively every week.

Satisfaction remained modest. Only 45% said they were satisfied with performance, and 42% reported issues with video calls and messaging platforms. Limited visibility was the most-cited challenge (51%), followed by dropped calls (42%) and integration problems with other enterprise systems (38%).

Observability standards

The survey also tracked adoption of OpenTelemetry, an open source framework for collecting and exporting telemetry data such as traces, metrics and logs. In manufacturing, 44% said they had fully implemented OpenTelemetry, and 42% said they were adopting it.

Respondents placed a strong emphasis on correlation across domains. Almost all (97%) agreed that cross-domain OpenTelemetry correlation is critical to their observability strategy. In addition, 93% said OpenTelemetry is a foundation for future initiatives such as AI-driven automation, while 37% said it is already mandated inside their organisation.

Networks and movement

Beyond tooling and standards, the study emphasised the role of data movement and network performance in AI readiness. Some 91% said the movement and sharing of data was important to their AI strategy, and 31% described it as critical and foundational to how they design and execute AI.

Plans for more centralised approaches are taking shape. Riverbed found that 75% of manufacturing respondents plan to establish an AI data repository strategy by 2028.

When asked about considerations for moving and scaling data, 96% cited network performance and capability. Cost of data movement and storage followed at 94%. AI model proximity to data and interoperability between environments both scored 93%.

Network security also featured prominently. In the survey, 79% said network performance and security were essential to their AI strategy, reflecting concerns about resilience, uptime, and the risks of moving sensitive production and operational data across sites and cloud environments.

The research was based on a global study of 1,200 business decision-makers, IT leaders and technical specialists across seven countries and multiple industries, including manufacturing. Coleman Parkes Research conducted the survey in July 2025.